The recent Assembly elections — the last major polling exercise before the 2019 Lok Sabha polls — were not devoid of Electronic Voting Machine (EVM) malfunctions. Though the discourse at present makes no distinction between a ‘malfunction’ (which suggests a technical defect) and ‘tampering’ (manipulation aimed at fraud), there were several reports of misbehaving EVMs. Alarmingly, in Madhya Pradesh alone, the number of votes polled did not match the number of votes counted in 204 out of the 230 constituencies. The Election Commission’s (EC) explanation is that the votes counted is the actual number of votes polled — a circular logic that precludes cross-verification.
A discrepancy of even one vote between votes polled and votes counted is unacceptable. This is not an unreasonably high standard but one followed by democracies worldwide. It might therefore be helpful to briefly look beyond the question that has hijacked the EVM debate — of how easy or tough it is to hack these machines — and consider the first principles of a free and fair election.
Electoral first principles
The reason a nation chooses to be a democracy is that it gives moral legitimacy to the government. The fount of this legitimacy is the people’s will. The people’s will is expressed through the vote, anonymously (the principle of secret ballot). Not only must this vote be recorded correctly and counted correctly, it must also be seen to be recorded correctly and counted correctly. The recording and counting process must be accessible to, and verifiable by, the public. So transparency, verifiability, and secrecy are the three pillars of a free and fair election.
Regardless of whether one is for or against EVMs, there is no getting away from the fact that any polling method must pass these three tests to claim legitimacy. Paper ballots obviously do. The voter can visually confirm that her selection has been registered, the voting happens in secret, and the counting happens in front of her representative’s eyes.
EVMs, however, fail on all three, as established by a definitive judgment of the German constitutional court in 2009. The court’s ruling forced the country to scrap EVMs and return to paper ballot. Other technologically advanced nations such as the Netherlands and Ireland have also abandoned EVMs.
If we take the first two criteria, EVMs are neither transparent nor verifiable. Neither can the voter see her vote being recorded, nor can it be verified later whether the vote was recorded correctly. What is verifiable is the total number of votes cast, not the choice expressed in each vote. An electronic display of the voter’s selection may not be the same as the vote stored electronically in the machine’s memory. This gap was why the Voter Verifiable Paper Audit Trail (VVPAT) was introduced.
Also read | An electoral intervention that has clicked
But VVPATs solve only one-half of the EVMs’ transparency/verifiability problem: the voting part. The counting part remains an opaque operation. If anyone suspects a counting error, there is no recourse, for an electronic recount is, by definition, absurd. Some believe the VVPATs can solve this problem too, through statistics.
At present, the EC’s VVPAT auditing is restricted to one randomly chosen polling booth per constituency. In a recent essay, K. Ashok Vardhan Shetty , a former IAS officer, demonstrates that this sample size will fail to detect faulty EVMs 98-99% of the time. He also shows that VVPATs can be an effective deterrent to fraud only on the condition that the detection of even one faulty EVM in a constituency must entail the VVPAT hand-counting of all the EVMs in that constituency. Without this proviso, VVPATs would merely provide the sheen of integrity without its substance.
The third criterion is secrecy. Here too, EVMs disappoint. With the paper ballot, the EC could mix ballot papers from different booths before counting, so that voting preferences could not be connected to a given locality. But with EVMs, we are back to booth-wise counting, which allows one to discern voting patterns and renders marginalised communities vulnerable to pressure. Totaliser machines can remedy this, but the EC has shown no intent to adopt them.
So, on all three counts — transparency, verifiability and secrecy — EVMs are flawed. VVPATs are not the answer either, given the sheer magnitude of the logistical challenges. The recent track record of EVMs indicates that the number of malfunctions in a national election will be high. For that very reason, the EC is unlikely to adopt a policy of hand-counting all EVMs in constituencies where faulty machines are reported, as this might entail hand-counting on a scale that defeats the very purpose of EVMs. And yet, this is a principle without which the use of VVPATs is meaningless.
Despite these issues, EVMs continue to enjoy the confidence of the EC, which insists that Indian EVMs, unlike the Western ones, are tamper-proof. But this is a matter of trust. Even if the software has been burnt into the microchip, neither the EC nor the voter knows for sure what software is running in a particular EVM. One has to simply trust the manufacturer and the EC. But as the German court observed, the precondition of this trust is the verifiability of election events, whereas in the case of EVMs, “the calculation of the election result is based on a calculation act which cannot be examined from outside”.
While it is true that the results come quicker and the process is cheaper with EVMs as compared to paper ballot, both these considerations are undeniably secondary to the integrity of the election. Another argument made in favour of the EVM is that it eliminates malpractices such as booth-capturing and ballot-box stuffing. In the age of the smartphone, however, the opportunity costs of ballot-box-stuffing and the risk of exposure are prohibitively high. In contrast, tampering with code could accomplish rigging on a scale unimaginable for booth-capturers. Moreover, it is nearly impossible to detect EVM-tampering. As a result, suspicions of tampering in the tallying of votes — as opposed to malfunction in registering the votes, which alone is detectable — are destined to remain in the realm of speculation. The absence of proven fraud might save the EVM for now, but its survival comes at a dangerous cost — the corrosion of people’s faith in the electoral process.
Yet there doesn’t have to be incontrovertible evidence of EVM-tampering for a nation to return to paper ballot. Suspicion is enough, and there is enough of it already. As the German court put it, “The democratic legitimacy of the election demands that the election events be controllable so that... unjustified suspicion can be refuted.” The phrase “unjustified suspicion” is pertinent. The EC has always maintained that suspicions against EVMs are unjustified. Clearly, the solution is not to dismiss EVM-sceptics as ignorant technophobes. Rather, the EC is obliged to provide the people of India a polling process capable of refuting unjustified suspicion, as this is a basic requirement for democratic legitimacy, not an optional accessory.